{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "complementary_code_covid.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "machine_shape": "hm"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S6TTn_99cZBU",
        "colab_type": "text"
      },
      "source": [
        "# The spread of COVID-19 conspiracy theories on social media and the effect of content moderation\n",
        "\n",
        "The study investigates the diffusion of conspiracy theories related to coronavirus between January and March 2020. Code by Orestis Papapkyriakopoulos (orestis.papakyriakopoulos@tum.de) and Juan Carlos Medina (juan.medina@tum.de).\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wFodi5sxiPoq",
        "colab_type": "text"
      },
      "source": [
        "**Define the path to the files:**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pSD7YN4QiTWI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "path = \"/content/drive/My Drive/covid_misinformation/data/\""
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ablbo5ImWfHI",
        "colab_type": "text"
      },
      "source": [
        "Necessary libraries:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k5066oXeeqwW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import warnings\n",
        "import pickle\n",
        "\n",
        "import matplotlib.colors as mcolors\n",
        "import matplotlib.ticker as ticker \n",
        "from mpl_toolkits.axes_grid1 import make_axes_locatable\n",
        "import numpy as np\n",
        "\n",
        "!pip install tick\n",
        "from tick.dataset import fetch_hawkes_bund_data\n",
        "from tick.hawkes import HawkesConditionalLaw\n",
        "from tick.plot import plot_hawkes_kernel_norms\n",
        "from tick.plot import plot_hawkes_kernels\n",
        "from tick.hawkes import HawkesSumGaussians, HawkesEM, HawkesSumExpKern, HawkesExpKern\n",
        "\n",
        "from google.colab import drive\n",
        "from google.colab import auth\n",
        "import gspread\n",
        "from oauth2client.client import GoogleCredentials"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cCWsNjBGVVY7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%matplotlib inline"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fcYZMm7KVcIk",
        "colab_type": "text"
      },
      "source": [
        "**Authenticate:**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3ymMMY3JbAxQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HWgn6ZRdV84Z",
        "colab_type": "text"
      },
      "source": [
        "# Coding"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1ypAHsjfV-ec",
        "colab_type": "text"
      },
      "source": [
        "We used the following coding for missinformation and source credibility:\n",
        "\n",
        "* **Missinformation**:\n",
        "\n",
        "\n",
        "0.   neutral or debunking content\n",
        "1.   content reinforcing conspiracy theories\n",
        "\n",
        "\n",
        "* **Source Credibility**:\n",
        "\n",
        "\n",
        "0.   mainstream\n",
        "1.   alternative\n",
        "2.   other\n",
        "\n",
        "* **Banning:**\n",
        "\n",
        "\n",
        "\n",
        "  *   **R** Removed\n",
        "  *   **F** Fact Check labeled\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y4RqWwcShr-z",
        "colab_type": "text"
      },
      "source": [
        "# Finding 1"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0WeuU3TEezAn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "with open(path + 'url_timestamps.pickle', 'rb') as handle:\n",
        "    df_multiplatform = pickle.load(handle)"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U3kbcxD6SbUs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_trajectories(list_of_platforms, regexp_misinformation, regexp_source):\n",
        "\n",
        "  id_lists = []\n",
        "\n",
        "  if \"twitter\" in list_of_platforms:\n",
        "    id_lists.append(3)\n",
        "\n",
        "  if \"facebook\" in list_of_platforms:\n",
        "    id_lists.append(4)\n",
        "\n",
        "  if \"reddit\" in list_of_platforms:\n",
        "    id_lists.append(5)\n",
        "\n",
        "  if \"4chan\" in list_of_platforms:\n",
        "    id_lists.append(6)\n",
        "\n",
        "  df_multiplatform_ =  df_multiplatform[df_multiplatform[df_multiplatform.iloc[:,id_lists] > 0].count(axis=1) > (len(id_lists) -1)]  \n",
        "  df_trajectories = df_multiplatform_[[\"misinformation\",\"source_credibility\",\"twitter_dt\", \"facebook_dt\", \"reddit_dt\", \"4chan_dt\"]][~df_multiplatform_[[\"twitter_order\", \"facebook_order\", \"reddit_order\", \"4chan_order\"]].isnull().all(1)]\n",
        "  df_trajectories_mis = df_trajectories[[\"twitter_dt\", \"facebook_dt\", \"reddit_dt\", \"4chan_dt\"]][df_trajectories[\"misinformation\"].str.contains(regexp_misinformation, na = False) & df_trajectories[\"source_credibility\"].str.contains(regexp_source, na = False)]\n",
        "  df_trajectories_mis[\"twitter_dt\"] = df_trajectories_mis[\"twitter_dt\"].apply(lambda d: np.asarray(d , dtype=np.double)if isinstance(d, list) else [])\n",
        "  df_trajectories_mis[\"facebook_dt\"] = df_trajectories_mis[\"facebook_dt\"].apply(lambda d: np.asarray(d , dtype=np.double) if isinstance(d, list) else [])\n",
        "  df_trajectories_mis[\"reddit_dt\"] = df_trajectories_mis[\"reddit_dt\"].apply(lambda d: np.asarray(d , dtype=np.double) if isinstance(d, list) else [])\n",
        "  df_trajectories_mis[\"4chan_dt\"] = df_trajectories_mis[\"4chan_dt\"].apply(lambda d: np.asarray(d , dtype=np.double) if isinstance(d, list) else [])\n",
        "\n",
        "  dt_lists = [name + \"_dt\" for name in list_of_platforms]\n",
        "\n",
        "  trajectories = df_trajectories_mis[dt_lists].values.tolist()\n",
        "\n",
        "  if \"twitter\" in list_of_platforms:\n",
        "    minimum = 1901000\n",
        "  else:\n",
        "    res = [min(l)  for q in trajectories for l in q]    \n",
        "    minimum = min(res)\n",
        "  trajectories_normalized = []\n",
        "  for url in trajectories:\n",
        "    list_of_platforms = []\n",
        "    for platform in url:\n",
        "      platform_ = platform[platform >= minimum] \n",
        "      plarform__ = (platform_ - minimum)/60\n",
        "      list_of_platforms.append(plarform__)\n",
        "    trajectories_normalized.append(list_of_platforms)\n",
        "\n",
        "  return(trajectories_normalized)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LZd9X7DYZxGp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "platform_list = [\"facebook\",\"twitter\",\"reddit\",\"4chan\"]\n",
        "\n",
        "def hawkes_process_calculation(misinformation_rgx):\n",
        "    platform_combinations = [(0,1), (0,2), (0,3), (1,2), (1,3), (2,3)]\n",
        "\n",
        "    amplitudes = np.arange(16.0).reshape(4,4)\n",
        "    amplitudes_combination = []\n",
        "    for combination in platform_combinations: \n",
        "          platform_c = [platform_list[combination[0]], platform_list[combination[1]]]\n",
        "          trajectories_all = get_trajectories(platform_c, regexp_misinformation = misinformation_rgx, regexp_source=\"0|1|2\" )\n",
        "          length = len(trajectories_all)\n",
        "          decays = [[0.000000001],[0.00000001],[0.0000001],[0.000001],[0.00001],[0.00005], [0.0001], [0.0002],[0.0003],[0.0004], [0.0005], [0.001], [0.002], [0.003], [0.004], [0.005], [0.006], [0.007], [0.008], [0.009], [0.01]]\n",
        "          c_values = [0.1]\n",
        "          scores = []\n",
        "          c_val = []\n",
        "          max_a = []\n",
        "          learners = []\n",
        "          for c_ in c_values: \n",
        "            for decay in decays:\n",
        "              learner = HawkesSumExpKern(decay, penalty='l2', C=c_)\n",
        "              learner.fit(trajectories_all[:int(0.8*length)])\n",
        "              score = learner.score(trajectories_all[int(0.8*length):])\n",
        "              if np.max(learner.adjacency)*decay[0] < decay[0] and np.max(learner.adjacency) <= 1:\n",
        "                scores.append(score)\n",
        "                c_val.append(c_)\n",
        "                max_a.append(np.max(learner.adjacency))\n",
        "                learners.append(learner)\n",
        "\n",
        "          best_decay = decays[scores.index(max(scores))]\n",
        "          best_c = c_val[scores.index(max(scores))]\n",
        "          max_a_ = max_a[scores.index(max(scores))]\n",
        "          best_learner = learners[scores.index(max(scores))]\n",
        "          amplitudes_combination.append((platform_c, best_learner.adjacency))\n",
        "          amplitudes[combination[0]][combination[0]] = best_learner.adjacency[0][0][0]\n",
        "          amplitudes[combination[0]][combination[1]] = best_learner.adjacency[0][1][0]\n",
        "          amplitudes[combination[1]][combination[0]] = best_learner.adjacency[1][0][0]\n",
        "          amplitudes[combination[1]][combination[1]] = best_learner.adjacency[1][1][0]\n",
        "\n",
        "          print(\"best_decay: \", best_decay, \"best_c: \", best_c, \" score : \", max(scores), \"max a: \", max_a_ )\n",
        "          print(scores)\n",
        "    return amplitudes"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eIdqW6edaxfR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%%capture capt\n",
        "amplitudes_non_conspiracy = hawkes_process_calculation(\"0\")\n",
        "amplitudes_conspiracy = hawkes_process_calculation(\"1\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fxecze3iSq7z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def make_colormap(seq):\n",
        "    \"\"\"Return a LinearSegmentedColormap\n",
        "    seq: a sequence of floats and RGB-tuples. The floats should be increasing\n",
        "    and in the interval (0,1).\n",
        "    \"\"\"\n",
        "    seq = [(None,) * 3, 0.0] + list(seq) + [1.0, (None,) * 3]\n",
        "    cdict = {'red': [], 'green': [], 'blue': []}\n",
        "    for i, item in enumerate(seq):\n",
        "        if isinstance(item, float):\n",
        "            r1, g1, b1 = seq[i - 1]\n",
        "            r2, g2, b2 = seq[i + 1]\n",
        "            cdict['red'].append([item, r1, r2])\n",
        "            cdict['green'].append([item, g1, g2])\n",
        "            cdict['blue'].append([item, b1, b2])\n",
        "    return mcolors.LinearSegmentedColormap('CustomMap', cdict)\n",
        "\n",
        "c = mcolors.ColorConverter().to_rgb\n",
        "rvb = make_colormap([c('white'),c('pink'), 0.33, c('pink'), 0.66, c('pink'), c('black')])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5QXob6d2SwdJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 482
        },
        "outputId": "e07e81f8-b128-4399-c7ad-59518ebc6fb7"
      },
      "source": [
        "pcolor_kwargs = {}\n",
        "pcolor_kwargs.setdefault(\"cmap\", rvb)\n",
        "max_amp = np.max(amplitudes_conspiracy)\n",
        "pcolor_kwargs.setdefault(\"vmin\", 0)\n",
        "pcolor_kwargs.setdefault(\"vmax\", max_amp)\n",
        "row_labels = ['${} \\\\rightarrow$'.format(i) for i in platform_list]\n",
        "column_labels = ['$\\\\rightarrow {}$'.format(i) for i in platform_list]\n",
        "x_label_alignment = 'center'\n",
        "rotate_x_labels = 0.\n",
        "\n",
        "fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(9, 6))\n",
        "fig.suptitle(\"Cross-platform conspiracy dynamics\", fontsize=18, x  = 0.9, y = 1.1)\n",
        "heatmap_1 = ax[0].pcolor(amplitudes_non_conspiracy, **pcolor_kwargs)\n",
        "\n",
        "# put the major ticks at the middle of each cell\n",
        "ax[0].set_xticks(np.arange(amplitudes_non_conspiracy.shape[0]) + 0.5, minor=False)\n",
        "ax[0].set_yticks(np.arange(amplitudes_non_conspiracy.shape[1]) + 0.5, minor=False)\n",
        "\n",
        "# want a more natural, table-like display\n",
        "ax[0].invert_yaxis()\n",
        "ax[0].xaxis.tick_top()\n",
        "\n",
        "ax[0].set_xticklabels(row_labels, minor=False, fontsize=17, \n",
        "                       rotation=rotate_x_labels, ha=x_label_alignment)\n",
        "ax[0].set_yticklabels(column_labels, minor=False, fontsize=17)\n",
        "\n",
        "\n",
        "heatmap_2 = ax[1].pcolor(amplitudes_conspiracy, **pcolor_kwargs)\n",
        "# put the major ticks at the middle of each cell\n",
        "ax[1].set_xticks(np.arange(amplitudes_conspiracy.shape[0]) + 0.5, minor=False)\n",
        "ax[1].set_yticks(np.arange(amplitudes_conspiracy.shape[1]) + 0.5, minor=False)\n",
        "\n",
        "# want a more natural, table-like display\n",
        "ax[1].invert_yaxis()\n",
        "ax[1].xaxis.tick_top()\n",
        "\n",
        "ax[1].set_xticklabels(row_labels, minor=False, fontsize=17, \n",
        "                       rotation=rotate_x_labels, ha=x_label_alignment)\n",
        "ax[1].set_yticklabels(column_labels, minor=False, fontsize=17)\n",
        "\n",
        "fig.subplots_adjust(right=1.6, wspace=0.3)\n",
        "divider = make_axes_locatable(ax[1])\n",
        "cax = divider.append_axes(\"right\", size=\"5%\", pad=0.5)\n",
        "fig.colorbar(heatmap_1, cax=cax)\n",
        "ax[0].set_title('neutral or debunking content', y=-0.1, fontsize=16)\n",
        "ax[1].set_title('content reinforcing conspiracy theories', y=-0.1, fontsize=16)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x432 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "go_ADkEUSzvc",
        "colab_type": "text"
      },
      "source": [
        "# Finding 2"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lXfjryoIR2NK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "outputId": "d544696c-1e3e-4455-9c24-630e567d4fa0"
      },
      "source": [
        "# buzzsumo only includes the reactions and shares for urls classified as misinformation. \n",
        "# Few urls did not appear in buzzsumo (not shared on Fb/twitter) and are missing here too\n",
        "df_buzzsumo = pd.read_csv(path + \"buzzsumo.csv\")\n",
        "df_buzzsumo.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>url_hash</th>\n",
              "      <th>facebook_total_reactions</th>\n",
              "      <th>twitter_shares</th>\n",
              "      <th>source_credibility</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>4</td>\n",
              "      <td>39</td>\n",
              "      <td>18</td>\n",
              "      <td>alternative</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>12</td>\n",
              "      <td>221</td>\n",
              "      <td>154</td>\n",
              "      <td>socialmedia</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>19</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>mainstream</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>20</td>\n",
              "      <td>47</td>\n",
              "      <td>9</td>\n",
              "      <td>mainstream</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>27</td>\n",
              "      <td>199</td>\n",
              "      <td>11000</td>\n",
              "      <td>alternative</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   url_hash  facebook_total_reactions  twitter_shares source_credibility\n",
              "0         4                        39              18        alternative\n",
              "1        12                       221             154        socialmedia\n",
              "2        19                         0               0         mainstream\n",
              "3        20                        47               9         mainstream\n",
              "4        27                       199           11000        alternative"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qAwby_NjAYzN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "warnings.filterwarnings(\"ignore\")\n",
        "\n",
        "def bar_plot_extended(x_data, y_data, x_data_dots, y_data_dots, title, xlabel, color_lines, color_dots):\n",
        "  fig = plt.figure(figsize=(5, 2))\n",
        "  ax = fig.add_axes([0, 0, 1, 1])\n",
        "  plt.hlines(y=y_data, xmin=0, xmax=x_data, color=color_lines, alpha=0.2, linewidth=20)\n",
        "\n",
        "  # create for each expense type a dot at the level of the expense percentage value\n",
        "  plt.plot(x_data_dots, y_data_dots, \"o\",  markersize=20, color=color_dots, alpha=1)\n",
        "  plt.title(title, fontsize=12, fontweight='black', color = '#333F4B')\n",
        "\n",
        "  # set labels style\n",
        "  ax.set_xlabel(xlabel, fontsize=12, fontweight='black', color = '#333F4B')\n",
        "  ax.set_ylabel('')\n",
        "  # change the style of the axis spines\n",
        "  ax.spines['top'].set_color('none')\n",
        "  ax.spines['right'].set_color('none')\n",
        "  ax.spines['left'].set_smart_bounds(True)\n",
        "  ax.spines['bottom'].set_smart_bounds(True)\n",
        "  ax.spines['bottom'].set_position(('axes', -0.1))\n",
        "  ax.spines['left'].set_position(('axes', 0.015))\n",
        "  ax.set_ylim([-1, 3])\n",
        "  plt.setp(ax.get_yticklabels(), fontweight=\"bold\")\n",
        "  plt.tight_layout() \n",
        "  plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J9RYeswPB_tL",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 239
        },
        "outputId": "32b1c46a-2d4f-4204-d45d-13d4fe1c23ec"
      },
      "source": [
        "mainstream_percentage = df_buzzsumo[df_buzzsumo.source_credibility == \"mainstream\"].shape[0] / df_buzzsumo.shape[0] * 100\n",
        "alternative_percentage = df_buzzsumo[df_buzzsumo.source_credibility == \"alternative\"].shape[0] / df_buzzsumo.shape[0] * 100\n",
        "socialmedia_percentage = df_buzzsumo[df_buzzsumo.source_credibility == \"socialmedia\"].shape[0] / df_buzzsumo.shape[0] * 100\n",
        "\n",
        "source = ['mainstream media', 'alternative media', 'social media & blogs']\n",
        "percent = [mainstream_percentage, alternative_percentage, socialmedia_percentage]\n",
        "\n",
        "bar_plot_extended(percent, source, percent, source, \"diffused content by source\", \"percentage %\", \"#e75480\", \"#8B0000\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x144 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6up9MJI0AoQu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 905
        },
        "outputId": "a3907e74-8847-4e82-dbeb-170034f59e2b"
      },
      "source": [
        "fb_median = df_buzzsumo.groupby(\"source_credibility\", as_index=False)[['facebook_total_reactions']].median()\n",
        "tw_median = df_buzzsumo.groupby(\"source_credibility\", as_index=False)[['twitter_shares']].median()\n",
        "fb_sum = df_buzzsumo.groupby(\"source_credibility\", as_index=False)[['facebook_total_reactions']].sum()\n",
        "tw_sum = df_buzzsumo.groupby(\"source_credibility\", as_index=False)[['twitter_shares']].sum()\n",
        "\n",
        "fb_median_mainstream = fb_median[fb_median.source_credibility == \"mainstream\"][\"facebook_total_reactions\"].to_list()[0]\n",
        "fb_median_alternative = fb_median[fb_median.source_credibility == \"alternative\"][\"facebook_total_reactions\"].to_list()[0]\n",
        "fb_median_socialmedia = fb_median[fb_median.source_credibility == \"socialmedia\"][\"facebook_total_reactions\"].to_list()[0]\n",
        "\n",
        "shares = [fb_median_mainstream, fb_median_alternative, fb_median_socialmedia]\n",
        "bar_plot_extended(shares, source, shares[0], source[0],\"median facebook reactions by source\", \"No. of reactions\", \"#007ACC\", \"#007ACC\")\n",
        "\n",
        "tw_median_mainstream = tw_median[tw_median.source_credibility == \"mainstream\"][\"twitter_shares\"].to_list()[0]\n",
        "tw_median_alternative = tw_median[tw_median.source_credibility == \"alternative\"][\"twitter_shares\"].to_list()[0]\n",
        "tw_median_socialmedia = tw_median[tw_median.source_credibility == \"socialmedia\"][\"twitter_shares\"].to_list()[0]\n",
        "\n",
        "shares = [tw_median_mainstream, tw_median_alternative, tw_median_socialmedia]\n",
        "bar_plot_extended(shares, source, shares, source,\"median twitter shares by source\", \"No. of shares\", \"#007ACC\", \"#007ACC\")\n",
        "\n",
        "fb_sum_mainstream = fb_sum[fb_sum.source_credibility == \"mainstream\"][\"facebook_total_reactions\"].to_list()[0]\n",
        "fb_sum_alternative = fb_sum[fb_sum.source_credibility == \"alternative\"][\"facebook_total_reactions\"].to_list()[0]\n",
        "fb_sum_socialmedia = fb_sum[fb_sum.source_credibility == \"socialmedia\"][\"facebook_total_reactions\"].to_list()[0]\n",
        "\n",
        "shares = [fb_sum_mainstream, fb_sum_alternative, fb_sum_socialmedia]\n",
        "bar_plot_extended(shares, source, shares[0:2], source[0:2],\"cumulative facebook reactions by source\", \"No. of reactions\", \"#007ACC\", \"#007ACC\")\n",
        "\n",
        "tw_sum_mainstream = tw_sum[tw_sum.source_credibility == \"mainstream\"][\"twitter_shares\"].to_list()[0]\n",
        "tw_sum_alternative = tw_sum[tw_sum.source_credibility == \"alternative\"][\"twitter_shares\"].to_list()[0]\n",
        "tw_sum_socialmedia = tw_sum[tw_sum.source_credibility == \"socialmedia\"][\"twitter_shares\"].to_list()[0]\n",
        "\n",
        "shares = [tw_sum_mainstream, tw_sum_alternative, tw_sum_socialmedia]\n",
        "bar_plot_extended(shares, source, shares[0:2], source[0:2],\"cumulative twitter shares by source\", \"No. of shares\", \"#007ACC\", \"#007ACC\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x144 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x144 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x144 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x144 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uwiD4EfCw6yu",
        "colab_type": "text"
      },
      "source": [
        "# Finding 3"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UKWwmdsET6tM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_trajectories_moderated(list_of_platforms, regexp_misinformation, regexp_source, moderated, platform):\n",
        "  \n",
        "  df_multiplatform_ = df_multiplatform\n",
        "\n",
        "  for pl in platform:\n",
        "    col_banned = pl + \"_banned\"\n",
        "    if moderated == True:\n",
        "        df_multiplatform_ = df_multiplatform_[df_multiplatform_[col_banned] > 0]\n",
        "    else:\n",
        "        df_multiplatform_ = df_multiplatform_[df_multiplatform_[col_banned] == 0]\n",
        "\n",
        "  id_lists = []\n",
        "\n",
        "  if \"twitter\" in list_of_platforms:\n",
        "    id_lists.append(3)\n",
        "\n",
        "  if \"facebook\" in list_of_platforms:\n",
        "    id_lists.append(4)\n",
        "\n",
        "  if \"reddit\" in list_of_platforms:\n",
        "    id_lists.append(5)\n",
        "\n",
        "  if \"4chan\" in list_of_platforms:\n",
        "    id_lists.append(6)\n",
        "\n",
        "  df_multiplatform_ =  df_multiplatform_[df_multiplatform_[df_multiplatform_.iloc[:,id_lists] > 0].count(axis=1) > (len(id_lists) -1)]  \n",
        "  df_trajectories = df_multiplatform_[[\"misinformation\",\"source_credibility\",\"twitter_dt\", \"facebook_dt\", \"reddit_dt\", \"4chan_dt\"]][~df_multiplatform_[[\"twitter_order\", \"facebook_order\", \"reddit_order\", \"4chan_order\"]].isnull().all(1)]\n",
        "  df_trajectories_mis = df_trajectories[[\"twitter_dt\", \"facebook_dt\", \"reddit_dt\", \"4chan_dt\"]][df_trajectories[\"misinformation\"].str.contains(regexp_misinformation, na = False) & df_trajectories[\"source_credibility\"].str.contains(regexp_source, na = False)]\n",
        "  df_trajectories_mis[\"twitter_dt\"] = df_trajectories_mis[\"twitter_dt\"].apply(lambda d: np.asarray(d , dtype=np.double)if isinstance(d, list) else [])\n",
        "  df_trajectories_mis[\"facebook_dt\"] = df_trajectories_mis[\"facebook_dt\"].apply(lambda d: np.asarray(d , dtype=np.double) if isinstance(d, list) else [])\n",
        "  df_trajectories_mis[\"reddit_dt\"] = df_trajectories_mis[\"reddit_dt\"].apply(lambda d: np.asarray(d , dtype=np.double) if isinstance(d, list) else [])\n",
        "  df_trajectories_mis[\"4chan_dt\"] = df_trajectories_mis[\"4chan_dt\"].apply(lambda d: np.asarray(d , dtype=np.double) if isinstance(d, list) else [])\n",
        "\n",
        "  dt_lists = [name + \"_dt\" for name in list_of_platforms]\n",
        "\n",
        "  trajectories = df_trajectories_mis[dt_lists].values.tolist()\n",
        "  if \"twitter\" in list_of_platforms:\n",
        "    minimum = 1901000\n",
        "  else:\n",
        "    res = [min(l)  for q in trajectories for l in q]    \n",
        "    minimum = min(res)\n",
        "  #print(minimum)\n",
        "  trajectories_normalized = []\n",
        "  for url in trajectories:\n",
        "    list_of_platforms = []\n",
        "    for platform in url:\n",
        "      platform_ = platform[platform >= minimum] \n",
        "      plarform__ = (platform_ - minimum)/60\n",
        "      list_of_platforms.append(plarform__)\n",
        "    trajectories_normalized.append(list_of_platforms)\n",
        "\n",
        "  return(trajectories_normalized)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "708DOldM_ELS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "platform_list = [\"twitter\",\"facebook\",\"reddit\",\"4chan\"]\n",
        "\n",
        "def hawkes_process_calculation_moderation(moderation, combination, plat):\n",
        "    trajectories_all = get_trajectories_moderated([platform_list[combination[0]],platform_list[combination[1]]], \"1\", \"0|1|2\", moderation, [platform_list[plat]])\n",
        "    length = len(trajectories_all)\n",
        "    print(length)\n",
        "    decays = [[0.000000001],[0.00000001],[0.0000001],[0.000001],[0.00001],[0.00005], [0.0001], [0.0002],[0.0003],[0.0004], [0.0005], [0.001], [0.002], [0.003], [0.004], [0.005], [0.006], [0.007], [0.008], [0.009], [0.01], [0.1]]\n",
        "    c_values = [0.1]\n",
        "    scores = []\n",
        "    c_val = []\n",
        "    max_a = []\n",
        "    learners = []\n",
        "    for c_ in c_values: \n",
        "      for decay in decays:\n",
        "        learner = HawkesSumExpKern(decay, penalty='l2', C=c_)\n",
        "        learner.fit(trajectories_all[:int(0.8*length)])\n",
        "        score = learner.score(trajectories_all[int(0.8*length):])\n",
        "        if np.max(learner.adjacency)*decay[0] < decay[0] and np.max(learner.adjacency) <= 1:\n",
        "          scores.append(score)\n",
        "          c_val.append(c_)\n",
        "          max_a.append(np.max(learner.adjacency))\n",
        "          learners.append(learner)\n",
        "  \n",
        "    best_decay = decays[scores.index(max(scores))]\n",
        "    best_c = c_val[scores.index(max(scores))]\n",
        "    max_a_ = max_a[scores.index(max(scores))]\n",
        "    best_learner = learners[scores.index(max(scores))]\n",
        "    return best_learner.adjacency, best_c"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G5znmZrx_3lE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%%capture\n",
        "platform_list = [\"twitter\",\"facebook\",\"reddit\",\"4chan\"]\n",
        "platform_combinations = [(0,1), (0,2), (0,3), (1,2), (1,3), (2,3)]\n",
        "\n",
        "amplitudes_combination = []\n",
        "\n",
        "for combination in platform_combinations:\n",
        "  for plat in combination:\n",
        "    # 4chan is missing\n",
        "    if plat == 3:\n",
        "        continue\n",
        "    moderated, best_c = hawkes_process_calculation_moderation(True, combination, plat)\n",
        "    unmoderated, best_c_un = hawkes_process_calculation_moderation(False, combination, plat)\n",
        "    amplitudes_combination.append(((platform_list[combination[0]],platform_list[combination[1]]), plat, moderated, unmoderated, best_c, best_c_un))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mZJTc63OUHOB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "pd.options.display.float_format = '{:,.2f}'.format\n",
        "indexes = [\"moderated\", \"unmoderated\", \"mod_platform\"]\n",
        "\n",
        "df_amplitudes = []\n",
        "for combination in amplitudes_combination:\n",
        "  amplitudes_array_moderated = combination[2].flatten()\n",
        "  amplitudes_array_unmoderated = combination[3].flatten()\n",
        "  names = [\"%s -> %s\" % (combination[0][0], combination[0][0]), \"%s -> %s\" % (combination[0][1], combination[0][0]), \"%s -> %s\" % (combination[0][0], combination[0][1]), \"%s -> %s\" % (combination[0][1], combination[0][1])]\n",
        "  df_temp = pd.DataFrame(np.vstack((amplitudes_array_moderated,amplitudes_array_unmoderated, [platform_list[combination[1]], platform_list[combination[1]],platform_list[combination[1]],platform_list[combination[1]] ])), index=indexes, columns=names)\n",
        "  df_amplitudes.append(df_temp)\n",
        "  "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CYzJUiCEUNx1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 990
        },
        "outputId": "3a0d1eb3-3715-41b2-d499-b3750925d0ac"
      },
      "source": [
        "df_platforms = pd.concat([x.T for x in df_amplitudes])\n",
        "df_platforms[\"moderated\"] = df_platforms[\"moderated\"].astype(float)\n",
        "df_platforms[\"unmoderated\"] = df_platforms[\"unmoderated\"].astype(float)\n",
        "\n",
        "df_platforms[\"p_change\"] = df_platforms.apply(lambda x: (x[0]- x[1])/x[1] if x[1] != 0.0 else 1, 1)\n",
        "df_platforms = df_platforms[~df_platforms.index.str.contains(\"^4chan\", na = False)]\n",
        "df_platforms[\"diff\"] = df_platforms[\"moderated\"] - df_platforms[\"unmoderated\"] \n",
        "df_platforms"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>moderated</th>\n",
              "      <th>unmoderated</th>\n",
              "      <th>mod_platform</th>\n",
              "      <th>p_change</th>\n",
              "      <th>diff</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; twitter</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.99</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-0.98</td>\n",
              "      <td>-0.97</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; twitter</th>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>twitter</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; facebook</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.81</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-0.98</td>\n",
              "      <td>-0.80</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; facebook</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.00</td>\n",
              "      <td>twitter</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; twitter</th>\n",
              "      <td>0.03</td>\n",
              "      <td>1.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>-0.97</td>\n",
              "      <td>-0.97</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; twitter</th>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>8.98</td>\n",
              "      <td>0.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; facebook</th>\n",
              "      <td>0.04</td>\n",
              "      <td>0.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; facebook</th>\n",
              "      <td>0.02</td>\n",
              "      <td>1.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>-0.98</td>\n",
              "      <td>-0.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; twitter</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.08</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-0.76</td>\n",
              "      <td>-0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; twitter</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.14</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-0.83</td>\n",
              "      <td>-0.11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-0.82</td>\n",
              "      <td>-0.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.02</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-0.84</td>\n",
              "      <td>-0.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; twitter</th>\n",
              "      <td>0.03</td>\n",
              "      <td>0.03</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.06</td>\n",
              "      <td>-0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; twitter</th>\n",
              "      <td>0.04</td>\n",
              "      <td>0.02</td>\n",
              "      <td>reddit</td>\n",
              "      <td>0.90</td>\n",
              "      <td>0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.02</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.86</td>\n",
              "      <td>-0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.02</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.81</td>\n",
              "      <td>-0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; twitter</th>\n",
              "      <td>1.00</td>\n",
              "      <td>0.09</td>\n",
              "      <td>twitter</td>\n",
              "      <td>10.36</td>\n",
              "      <td>0.91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>twitter -&gt; 4chan</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>twitter</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; facebook</th>\n",
              "      <td>0.05</td>\n",
              "      <td>0.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; facebook</th>\n",
              "      <td>0.06</td>\n",
              "      <td>0.37</td>\n",
              "      <td>facebook</td>\n",
              "      <td>-0.84</td>\n",
              "      <td>-0.31</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.98</td>\n",
              "      <td>facebook</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-0.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; facebook</th>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; facebook</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.53</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.96</td>\n",
              "      <td>-0.51</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>reddit</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; reddit</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.45</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.99</td>\n",
              "      <td>-0.44</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; facebook</th>\n",
              "      <td>0.02</td>\n",
              "      <td>1.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>-0.98</td>\n",
              "      <td>-0.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>facebook -&gt; 4chan</th>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>facebook</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; reddit</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.46</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.95</td>\n",
              "      <td>-0.44</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>reddit -&gt; 4chan</th>\n",
              "      <td>0.02</td>\n",
              "      <td>0.65</td>\n",
              "      <td>reddit</td>\n",
              "      <td>-0.96</td>\n",
              "      <td>-0.63</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                      moderated  unmoderated mod_platform  p_change  diff\n",
              "twitter -> twitter         0.02         0.99      twitter     -0.98 -0.97\n",
              "facebook -> twitter        0.01         0.00      twitter      1.00  0.01\n",
              "twitter -> facebook        0.02         0.81      twitter     -0.98 -0.80\n",
              "facebook -> facebook       0.02         0.00      twitter      1.00  0.02\n",
              "twitter -> twitter         0.03         1.00     facebook     -0.97 -0.97\n",
              "facebook -> twitter        0.01         0.00     facebook      8.98  0.01\n",
              "twitter -> facebook        0.04         0.00     facebook      1.00  0.04\n",
              "facebook -> facebook       0.02         1.00     facebook     -0.98 -0.98\n",
              "twitter -> twitter         0.02         0.08      twitter     -0.76 -0.06\n",
              "reddit -> twitter          0.02         0.14      twitter     -0.83 -0.11\n",
              "twitter -> reddit          0.00         0.01      twitter     -0.82 -0.01\n",
              "reddit -> reddit           0.00         0.02      twitter     -0.84 -0.01\n",
              "twitter -> twitter         0.03         0.03       reddit     -0.06 -0.00\n",
              "reddit -> twitter          0.04         0.02       reddit      0.90  0.02\n",
              "twitter -> reddit          0.00         0.02       reddit     -0.86 -0.02\n",
              "reddit -> reddit           0.00         0.02       reddit     -0.81 -0.02\n",
              "twitter -> twitter         1.00         0.09      twitter     10.36  0.91\n",
              "twitter -> 4chan           0.00         0.00      twitter     -1.00 -0.00\n",
              "facebook -> facebook       0.05         0.00     facebook      1.00  0.05\n",
              "reddit -> facebook         0.06         0.37     facebook     -0.84 -0.31\n",
              "facebook -> reddit         0.00         0.00     facebook      1.00  0.00\n",
              "reddit -> reddit           0.00         0.98     facebook     -1.00 -0.98\n",
              "facebook -> facebook       0.00         1.00       reddit     -1.00 -1.00\n",
              "reddit -> facebook         0.02         0.53       reddit     -0.96 -0.51\n",
              "facebook -> reddit         0.00         0.00       reddit      1.00  0.00\n",
              "reddit -> reddit           0.00         0.45       reddit     -0.99 -0.44\n",
              "facebook -> facebook       0.02         1.00     facebook     -0.98 -0.98\n",
              "facebook -> 4chan          0.00         1.00     facebook     -1.00 -1.00\n",
              "reddit -> reddit           0.02         0.46       reddit     -0.95 -0.44\n",
              "reddit -> 4chan            0.02         0.65       reddit     -0.96 -0.63"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jr3wzDzoKTdO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 127
        },
        "outputId": "160f06e3-c88d-43f7-c7d8-2d894d8a88b7"
      },
      "source": [
        "df_platforms[\"from\"], df_platforms[\"to\"] = zip(*df_platforms.index.str.split(\" -> \").tolist())\n",
        "print(\"In-platform effects\")\n",
        "df_in = df_platforms[df_platforms.apply(lambda x: x[\"from\"]==x[\"to\"]==x[\"mod_platform\"], 1)]\n",
        "df_in.groupby(\"mod_platform\").mean()[\"diff\"]/df_in.groupby(\"mod_platform\").mean()[\"unmoderated\"]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "In-platform effects\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "mod_platform\n",
              "facebook   -0.96\n",
              "reddit     -0.97\n",
              "twitter    -0.10\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HB5I3bTEKqgw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 127
        },
        "outputId": "8c856595-77bf-477a-d36c-8c422762e497"
      },
      "source": [
        "print(\"Cross-platform effects\")\n",
        "df_out = df_platforms[df_platforms.apply(lambda x: (x[\"from\"]!=x[\"to\"]) & (x[\"from\"]==x[\"mod_platform\"]), 1)]\n",
        "df_out.groupby(\"mod_platform\").mean()[\"diff\"]/df_out.groupby(\"mod_platform\").mean()[\"unmoderated\"]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Cross-platform effects\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "mod_platform\n",
              "facebook   -0.99\n",
              "reddit     -0.93\n",
              "twitter    -0.98\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3skMf9BlKzDG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 90
        },
        "outputId": "1322ee3e-dfc4-481d-884f-e33374a9c6f5"
      },
      "source": [
        "print(\"Overall effects\")\n",
        "print(\"facebook\", ( df_platforms[df_platforms.index.str.contains(\"^facebook\", na = False)][\"moderated\"].mean()  - df_platforms[df_platforms.index.str.contains(\"^facebook\", na = False)][\"unmoderated\"].mean() )/df_platforms[df_platforms.index.str.contains(\"^facebook\", na = False)][\"unmoderated\"].mean())\n",
        "print(\"reddit\", ( df_platforms[df_platforms.index.str.contains(\"^twitter\", na = False)][\"moderated\"].mean()  - df_platforms[df_platforms.index.str.contains(\"^twitter\", na = False)][\"unmoderated\"].mean() )/df_platforms[df_platforms.index.str.contains(\"^twitter\", na = False)][\"unmoderated\"].mean())\n",
        "print(\"twitter\", ( df_platforms[df_platforms.index.str.contains(\"^reddit\", na = False)][\"moderated\"].mean()  - df_platforms[df_platforms.index.str.contains(\"^reddit\", na = False)][\"unmoderated\"].mean() )/df_platforms[df_platforms.index.str.contains(\"^reddit\", na = False)][\"unmoderated\"].mean())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Overall effects\n",
            "facebook -0.9691567216256533\n",
            "reddit -0.6195335425705961\n",
            "twitter -0.9442188556608188\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aVfvFajZF_mQ",
        "colab_type": "text"
      },
      "source": [
        "# Finding 4"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JlQs85itdGkA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "twitter_banned = pd.read_csv(path + \"twitter_banned.csv\")\n",
        "facebook_banned = pd.read_csv(path + \"facebook_banned.csv\")\n",
        "reddit_banned = pd.read_csv(path + \"reddit_banned.csv\")\n",
        "youtube_banned = pd.read_csv(path + \"youtube_banned.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SsPhbUWJdHSX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "twitter_march = twitter_banned[twitter_banned[\"banned_march\"].isin([\"R\", \"F\"])].shape[0] / twitter_banned.shape[0] * 100\n",
        "twitter_april = twitter_banned[twitter_banned[\"banned_april\"].isin([\"R\", \"F\"])].shape[0] / twitter_banned.shape[0] * 100\n",
        "facebook_march = facebook_banned[facebook_banned[\"banned_march\"].isin([\"R\", \"F\"])].shape[0] / facebook_banned.shape[0] * 100\n",
        "facebook_april = facebook_banned[facebook_banned[\"banned_april\"].isin([\"R\", \"F\"])].shape[0] / facebook_banned.shape[0] * 100\n",
        "reddit_march = reddit_banned[reddit_banned[\"banned_march\"].isin([\"R\", \"F\"])].shape[0] / reddit_banned.shape[0] * 100\n",
        "reddit_april = reddit_banned[reddit_banned[\"banned_april\"].isin([\"R\", \"F\"])].shape[0] / reddit_banned.shape[0] * 100\n",
        "youtube_march = youtube_banned[youtube_banned[\"banned_march\"].isin([\"R\", \"F\"])].shape[0] / youtube_banned.shape[0] * 100\n",
        "youtube_april = youtube_banned[youtube_banned[\"banned_april\"].isin([\"R\", \"F\"])].shape[0] / youtube_banned.shape[0] * 100"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H2UsE08cOMnT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 369
        },
        "outputId": "128b8de1-33c3-4358-b3ff-e8c0c6704980"
      },
      "source": [
        "fig = plt.figure(figsize=(5, 4))\n",
        "ax = fig.add_axes([0, 0, 1, 1])\n",
        "source = [0, 0.6, 1.5, 2.1, 3, 3.6, 4.5, 5.1]\n",
        "percent = [youtube_april, youtube_march, facebook_april, facebook_march, twitter_april, twitter_march, reddit_april, reddit_march]\n",
        "plt.hlines(y=source, xmin=0, xmax=percent,colors=[\"#e75480\", '#007ACC', \"#e75480\", '#007ACC', \"#e75480\", '#007ACC', \"#e75480\", '#007ACC'], alpha=0.2, linewidth=20)\n",
        "plt.yticks([0.2, 1.7, 3.2, 4.7],['youtube', 'facebook', 'twitter', 'reddit'])\n",
        "\n",
        "\n",
        "# create for each expense type a dot at the level of the expense percentage value\n",
        "colors = ['#8B0000', '#007ACC', '#8B0000', '#007ACC', '#8B0000', '#007ACC', '#8B0000', '#007ACC']\n",
        "for i,_ in enumerate(percent):\n",
        "  plt.plot(percent[i], source[i], \"o\",color=colors[i], markersize=20, alpha=1)\n",
        "\n",
        "import matplotlib.patches as mpatches\n",
        "\n",
        "red_patch = mpatches.Patch(color='#e75480', alpha=0.2, label='May 2020')\n",
        "blue_patch = mpatches.Patch(color='#007ACC', alpha=0.2, label='April 2020')\n",
        "\n",
        "plt.legend(handles=[blue_patch, red_patch], title=\"checked in\", loc=4, fancybox=True)\n",
        "\n",
        "plt.title(\"proportion of banned content by platform\", fontsize=12, fontweight='black', color = '#333F4B')\n",
        "\n",
        "# set labels style\n",
        "ax.set_xlabel('Percentage', fontsize=12, fontweight='black', color = '#333F4B')\n",
        "ax.set_ylabel('')\n",
        "# change the style of the axis spines\n",
        "ax.spines['top'].set_color('none')\n",
        "ax.spines['right'].set_color('none')\n",
        "ax.spines['left'].set_smart_bounds(True)\n",
        "ax.spines['bottom'].set_smart_bounds(True)\n",
        "ax.spines['bottom'].set_position(('axes', 0.0))\n",
        "ax.spines['left'].set_position(('axes', 0.015))\n",
        "ax.set_ylim([-1,6])\n",
        "plt.setp(ax.get_yticklabels(), fontweight=\"bold\")\n",
        "plt.tight_layout() \n",
        "plt.show()\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TNShbYOVYlDK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 369
        },
        "outputId": "10dec35e-0987-4a63-bae4-2372f3fa9293"
      },
      "source": [
        "fig = plt.figure(figsize=(5,4))\n",
        "ax = fig.add_axes([0,0,1,1])\n",
        "source = [0,0.6,1.2, 2,2.6,3.2, 4,4.6,5.2]\n",
        "percent = [4.617117117117117, 13.967258794845003, 6.6543438077634, 36.27535143255714, 45.72227510709186, 47.68133174791914, 23.809523809523808, 28.57142857142857, 44.742729306487694]\n",
        "\n",
        "\n",
        "plt.hlines(y=source, xmin=0, xmax=percent,colors=['#007ACC',\"#e75480\",\"#000000\",'#007ACC',\"#e75480\",\"#000000\",'#007ACC',\"#e75480\",\"#000000\"], alpha=0.1, linewidth=20)\n",
        "plt.yticks([0.6,2.6,4.6],['facebook', 'twitter', 'reddit'])\n",
        "\n",
        "\n",
        "# create for each expense type a dot at the level of the expense percentage value\n",
        "colors = ['#007ACC','#8B0000','#000000','#007ACC','#8B0000','#000000','#007ACC','#8B0000','#000000']\n",
        "for i,_ in enumerate(percent):\n",
        "  plt.plot(percent[i], source[i], \"o\",color=colors[i], markersize=20, alpha=1)\n",
        "\n",
        "import matplotlib.patches as mpatches\n",
        "\n",
        "black_patch = mpatches.Patch(color='#000000', alpha=0.2, label='social media & blogs')\n",
        "red_patch = mpatches.Patch(color='#e75480', alpha=0.2, label='alternative media')\n",
        "blue_patch = mpatches.Patch(color='#007ACC', alpha=0.2, label='mainstream media')\n",
        "\n",
        "plt.legend(handles=[black_patch, red_patch,blue_patch], loc=4, fancybox=True)\n",
        "\n",
        "#ax.bar(source,percent)\n",
        "plt.title(\"source of banned content by platform\", fontsize=12, fontweight='black', color = '#333F4B')\n",
        "\n",
        "# set labels style\n",
        "ax.set_xlabel('Relative percentage', fontsize=12, fontweight='black', color = '#333F4B')\n",
        "ax.set_ylabel('')\n",
        "# change the style of the axis spines\n",
        "ax.spines['top'].set_color('none')\n",
        "ax.spines['right'].set_color('none')\n",
        "ax.spines['left'].set_smart_bounds(True)\n",
        "ax.spines['bottom'].set_smart_bounds(True)\n",
        "ax.spines['bottom'].set_position(('axes', 0.0))\n",
        "ax.spines['left'].set_position(('axes', 0.015))\n",
        "ax.set_ylim([-1,6])\n",
        "plt.setp(ax.get_yticklabels(), fontweight=\"bold\")\n",
        "plt.tight_layout() \n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 360x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}